The information in this article originally appeared in an Equip Academy presentation. Watch the presentation here, and register for future Equip Academy events to learn about other eating disorder-related topics and earn free CE credits.

When most people think about eating disorders, they think of anorexia, bulimia, or maybe binge eating disorder. But it’s actually a lesser-known eating disorder that has the largest financial impact in the United States. In 2018-2019, other specified feeding and eating disorder (OSFED) cost the country $22.8 billion, or 35% of the $64.7 billion total financial burden of eating disorders. For reference, binge eating disorder cost the country $19.4 billion, while bulimia and anorexia cost $11.4 billion and $11.2 billion, respectively. OSFED is just as serious as other eating disorders, as evidenced by its massive economic toll.

So what is OSFED? Put most simply, it’s a catch-all diagnosis used to describe eating disorders that significantly impact a person’s health and life but do not meet diagnostic criteria for another eating disorder. Though it can be used as a general diagnosis, there are several specific diagnosis that fall under the OSFED umbrella, including:

• Binge eating disorder of low frequency or limited duration
• Bulimia nervosa of low frequency or limited duration
• Night eating syndrome
• Purging disorder
• Atypical anorexia (AAN)

Here, we’ll be looking closely at one of the most misunderstood OSFED diagnoses, atypical anorexia.

What is atypical anorexia?

Atypical anorexia is a diagnosis that describes someone who meets all of the criteria for anorexia nervosa (AN) except that despite significant weight loss, the individual’s weight is within or above the normal range.

Atypical anorexia and anorexia nervosa share many of the same diagnostic criteria, including:

• Restriction of energy intake relative to requirements
• Intense fear of gaining weight or becoming fat, or persistent behavior that interferes with weight gain
• Disturbance in the way in which one’s body weight or shape is experienced, undue influence of body weight or shape on self-evaluation, or persistent lack of recognition of the seriousness of illness

The primary difference between the two is that in anorexia nervosa, the restricted intake leads to significantly low body weight, whereas in atypical anorexia nervosa, it does not. Additionally, those with anorexia nervosa fail to recognize the seriousness of their illness and low body weight, whereas those with atypical anorexia fail to recognize the seriousness of their illness only (since low body weight isn’t present).

Despite the name, atypical anorexia is actually more “typical” than anorexia nervosa: AAN has a lifetime prevalence of 4.9%, as compared to 0.6% for AN. And this number is likely low, given that many people with atypical anorexia are misdiagnosed or undiagnosed due to weight stigma and anti-fat bias in medical spaces. Often, people with atypical anorexia are assumed to have bulimia or binge eating disorder based on their body size alone (if an eating disorder is suspected at all). What’s more, patients with atypical anorexia may even be prescribed restrictive eating behaviors, which can exacerbate symptoms and highlights the confusion that many patients with AAN experience.

Because of stigma, misconceptions, and lack of awareness, there’s currently limited research on the effective treatment of atypical anorexia. In fact, atypical anorexia wasn’t even added to the DSM until the publication of the DSM-5 in 2013.

Prior to 2013, patients with atypical anorexia presented for treatment, but were often turned away or given a diagnosis of eating disorder not otherwise specified (EDNOS). In an effort to reduce the number of patients diagnosed with EDNOS, atypical anorexia was added to the DSM-5 in 2013 under the new OSFED diagnosis. It’s important to note that patients who met AAN criteria were always there, but were often overlooked by doctors for treatment because there was no established diagnosis that fit their presentation.

Is atypical anorexia less serious than anorexia nervosa?

The atypical anorexia diagnosis has been referred to as “subthreshold anorexia nervosa” by some researchers, likening it to other OSFED categories of bulimia nervosa of low frequency or duration or BED of low frequency or duration. However, this separation puts patients in danger, because data show clearly that atypical anorexia is just as serious as anorexia nervosa.

Patients with atypical anorexia experience:

• More severe eating disorder symptoms and greater body dissatisfaction
• Similar rates of suicidality and suicide attempts to those with anorexia nervosa
• High levels of depression and anxiety, similar to rates among those with anorexia nervosa

Atypical anorexia is also more prevalent among cisgender boys/men, BIPOC individuals, and transgender and gender diverse people, all of whom already experience barriers to care.

Atypical anorexia comes with many of the same physical health complications as anorexia nervosa, including:

• Multiple electrolyte imbalances
• Risk of refeeding syndrome
• Impaired renal function
• Bradycardia
• Hypotension
• Delayed gastric emptying
• Impaired liver function
• Anemia
• Amenorrhea

How do you actually differentiate between AAN and AN?

That’s a good question, and we wondered the same thing. To get to the answer, we began by trying to gain clarity about the difference between atypical anorexia and anorexia nervosa.

The primary distinction between AN and AAN is that people with AAN are at a weight that is within or above normal range—but there’s no consensus on what this means. In fact, a recent systematic review found that in 75 separate studies of AAN, 29 unique definitions of AAN were used.

The four most common definitions for atypical anorexia are:

1. AAN-DSM5: BMI at admission to treatment is greater than or equal to19 kg/m2 for adults aged 18 or older, or above the median BMI for patients under 18.
2. AAN-BMI-18.5: BMI at admission to treatment is greater than 18.5 kg/m2 for adults or over 85% of the median BMI for adolescents and children.
3. AAN-BMI-10th Percentile: BMI at admission to treatment is over the 10th percentile (only relevant for children/adolescents).
4. AAN-DSM5 HLOC: BMI at admission to treatment is greater than or equal to 19 kg/m2 for adults or above the median for adolescents, and they had reported no previous treatment experience beyond outpatient treatment. The BMI thresholds are the same as the DSM-5 definition. However, patients with previous eating disorder treatment at a higher level of care (e.g., treatment at a residential treatment center) were classified as AN even if their BMI was above the threshold, as previous treatment may have resulted in weight gain that would place patient BMI above the threshold.21–23

What does treatment for AAN look like?

Little information is available about treatment for atypical anorexia, and few studies have examined effectiveness. Some research supports the preliminary effectiveness of family-based treatment (FBT) and PHP-level care, but this research notes complicating factors related to treatment.

A case series examining FBT for AAN described treatment and outcomes for 42 adolescents with atypical anorexia. The research showed symptom improvement and preliminary support for using FBT to treat adolescents with atypical anorexia, but called for more in-depth studies into its effectiveness.

Providers who have used FBT to treat atypical anorexia report a few challenges, most notably:

• Difficulty identifying when an AAN diagnosis is appropriate
• Pushback in setting target weights for adolescents with AAN
• Increased need to emphasize seriousness of AAN to families during FBT

Equip research on AAN treatment

To learn more about what treatment works for AAN—and whether AAN requires a different treatment approach than AN—we conducted a robust study that compared treatment for the two populations. Our aims with this research were to:

• Compare symptom severity at admission between AN and AAN
• Evaluate treatment response among AAN patients
• Compare change over treatment among patients with AN and AAN

We used all four of the most common definitions of AAN to ensure that outcomes were not affected by the definition being used.

Our patient population included 1,463 patients with AN or OSFED, with a median age of 15.3. Seventy percent were white, and 85.4% were cisgender girls or women. The outcomes we measured were eating disorder symptoms, anxiety, depression, caregiver burden, and weight.

The number of patients classified into the AN and AAN analysis groups varied greatly depending on which definition of AAN was used. Some patients were classified as AAN under all four definitions (n = 308), some were classified as AN under all four definitions (n = 299), and a majority were mixed (n = 856).

Of the patients classified as AAN, 70.4%-80.7% required weight restoration, compared with 85% to 94% for those with AN, depending on the definition used. Eating disorder symptoms were significantly greater for those with atypical anorexia compared with anorexia for three out of the four AAN definitions. For patients with AAN who required weight restoration, the amount of weight gain required varied depending on the definition used. BMI-10th AAN required the most weight gain to meet target weight (an average of 17.1 lbs).

We had three key takeaways from symptoms at admission:

• Most patients with AAN required weight restoration, regardless of how they were classified
• Patients with AAN had less weight to gain to meet their target weights, but the average weight to gain was 14+ lbs no matter how we defined AAN
• Patients with AAN reported more severe eating disorder symptoms than those with AN

Upon beginning treatment, there were similarities and differences between patients with AAN and AN. All patients improved with treatment on all measures, and there was no difference in rate of improvement. However, weight gain progressed more slowly for patients with AAN.

Since the cut-off threshold between AN and AAN did not have substantial impact on outcomes between definitions, we tested BMI as a continuous variable. If indeed there is a “true” distinction between AN and AAN based on weight status that warrants separate diagnoses, we would expect symptom presentation and outcomes to be associated with BMI. There was no significant association between BMI and symptom presentation or treatment outcomes.

Our research showed that there were no differences in treatment response between groups no matter how we defined AAN. FBT created meaningful symptom improvement independent of diagnostic classification or BMI. We suggest that weight status in AN and AAN is, similarly, not a factor requiring diagnostic distinction. Rather, weight status may provide information about the individual’s experience in the world and may be associated with experiences of clinical focus, such as widespread weight stigma, delayed access to care, and praise for weight loss associated with one’s eating disorder.

Explore future Equip Academy eventsEarn free CE credits and learn more about eating disorders from Equip's clinical experts.

See upcoming events

Implications for practice

Although more research is needed to better understand atypical anorexia, how it differs from anorexia nervosa, and how best to treat it, our research does provide some clear implications for treatment of AAN at this point.

Considerations for providers treating AAN include:

• Effectiveness of our treatment wasn’t dependent on classification into AN versus AAN, or BMI
• There’s growing evidence that AN and AAN may be treated in the same or similar ways
• Comprehensive Health At Every Size (HAES) training for providers and HAES-informed practice may especially benefit patients with AAN

Despite the fact that our study showed no significant differences between treatment effectiveness for patients with AAN vs AN, there are some unique factors that apply to treatment for patients with AAN. Although it’s not diagnostically important, weight may impact patients’ experience with their eating disorder, as patients at higher weights often experience the compounding effects of weight stigma.

This can manifest in a number of different ways throughout the trajectory of their illness.

During the development of the eating disorder

• Provider pathologizes body and weight
• Dual messaging from family and medical community
• Provider-recommended weight loss

Pre-treatment

• Provider may not notice eating disorder
• Missed screening opportunities
• Provider discounts or minimizes eating disorder
• Provider encourages or congratulations eating disordered behaviors
• Delayed diagnosis and/or limited access to care

During treatment

• Provider minimizes or denies eating disorder
• Misdiagnosis and missed symptoms
• Provider encourages eating disorder
• Weight stigma in higher levels of eating disorder care
• Systemic issues of weight bias
• Reluctance by caregivers or support people to engage with or support treatment
• Increased fear or grief related to weight gain in recovery

Post-treatment

• Provider now acknowledging eating disorder
• Weight/weight loss-focused care
• Aftercare may trigger relapse
• Overt weight discrimination
• Systemic issues of weight bias

Working with patients with AAN and their families means being aware of these unique experiences and being able to address them. It also means paying attention to other identities that may intersect with body size and impact patients’ experiences, and incorporating these factors into treatment, relapse prevention, and recovery planning.

From our research, we make three recommendations:

1. AAN should not be a separate diagnosis from AN.
2. Current diagnostic codes are not clinically useful and perpetuate fatphobia.
3. AN should be an appropriate diagnosis no matter weight, BMI, or body size.

Atypical anorexia is a very serious eating disorder, but it is also fully treatable—as long as providers learn how to recognize it and take the above considerations into account during treatment.

Schedule a call with our team to refer a patient or learn more about Equip’s evidence-based eating disorder treatment for atypical anorexia.